Charles Lyons-Burt
Government agencies are increasingly turning to artificial intelligence to modernize operations, and they’re doing so amid a growing set of standards, including the Biden administration’s 2022 AI Bill of Rights blueprint and the 2023 Executive Order for AI. Most recently, the Bipartisan Senate AI Working Group introduced a roadmap for AI policy in the Senate in May 2024, designed to drive AI innovation in the U.S.
As adoption continues, government technology teams are working to evolve best practices and identify the most impactful applications for the technology, all while seeking to minimize risk and remain compliant with enhanced regulatory guidance around AI.
Mike Adams, program executive for AI solutions at Carahsoft, recently sat down with ExecutiveBiz for a discussion on how federal, state and local government and higher education information technology leaders are balancing these issues as they work to implement AI technology for more efficient, secure and service-oriented operations.
Can you tell us about your role at Carahsoft as head of the company’s AI vertical?
Carahsoft connects agencies with technology providers across numerous focus areas including cybersecurity and the Cybersecurity Maturity Model Certification, multi-cloud, DevSecOps, autonomy and robotics, and more. Given that AI is a prevalent force across these fronts, and as its own standalone technology, my role is to bolster our portfolio of AI solutions available to government customers, while also helping to streamline procurement through a host of contract vehicles.
Our company is committed to building stronger public and private partnerships through knowledge-sharing. Altogether, my goal is to simplify the roles of government procurement teams and the systems integrators who own the mission requirements and must deploy and maintain the AI systems that operate within government agencies.
Carahsoft recently held its inaugural AI for Government Summit. Can you share thoughts about what went into the summit and some of the key takeaways from panels throughout the day?
The half-day event brought together 20 Carahsoft vendor partners and dozens of government leaders to discuss the adoption of AI in government, with a focus on several core areas of the White House Executive Order such as AI infrastructure in the data center and cloud, cybersecurity, and generative AI.
A main theme of the summit was “Taking the Lead in a New Era,” with panels devoted to best practices for deploying AI safely, securely and effectively. Much of the conversation was focused around getting data AI-ready through more granular, secure and agile data management; and getting infrastructure AI-ready with modernization rounds aimed at reducing reliance on outmoded legacy systems.
The key takeaway throughout each session was that AI is a tremendous resource for government, provided it is designed strategically and deployed responsibly. This theme was first introduced in an engaging keynote address by Department of Homeland Security Undersecretary for Science and Technology Dr. Dimitri Kusnezov, who provided powerful examples of AI’s great potential in government operations.
What are some of the most exciting and impactful AI use cases you’re seeing lately in government?
There is a pressing need for AI in cybersecurity to better protect agency operations at scale with automated alert management, stronger threat monitoring and faster response to attacks. One reason this is such a priority is that malicious actors are also using AI for more sophisticated and large-scale attacks. Government cybersecurity teams must embrace AI to match the power of the threat.
AI can also monitor for fraud to ensure stronger tax compliance or accurate distribution of government benefits. Where there is fraud, AI can speed detection and proactively flag suspicious behavioral patterns that would be difficult and time consuming to spot manually. These are also great examples of how the investment in AI typically pays for itself over time.
AI has promising applications in connecting citizens with government services; this includes contact centers that automate certain client interactions and platforms that case managers use to streamline processes for administering benefits. AI can also help optimize staffing and operations at VA hospitals, analyze weather patterns to predict and respond to natural disasters and provide supply chain or battlefield decision support for military operations.
Can you discuss generative AI in particular? How can government agencies ideally leverage this new form of AI that has become more popular since the release of ChatGPT?
Generative AI is of tremendous interest to agencies as they see the power of large language models to enhance numerous government use cases, such as productivity, automation of daily or administrative tasks and processes like acquisition and contract management.
However, this form of AI is subject to the same responsible use caveat that applies to all AI. Generative AI results can contain errors and insufficient domain context for an agency’s particular mission objective. As a result, we are connecting government IT leaders with solution partners who can deliver generative AI with the right governance around security, access and data quality that agencies require.
How do you choose your partners and ensure they have the right credentials and solutions to help agencies leverage AI?
A lot of due diligence goes into understanding the many different AI sub-verticals and use cases, and then evaluating the most appropriate resources to match government needs. We vet a company not just on whether they have a good product, but how well-suited that product is for the most pressing goals that government agencies are seeking to achieve in their AI missions.
We start by assessing vendors’ underlying technology and processes. This goes beyond just looking at FedRAMP and other certifications to also examine whether they use agile data management techniques or whether they can support accelerated computing for AI training and processing.
We examine all of these things and more: Who are the people there? What kind of investments are they making? Are they resilient? These are just some of the questions we ask and the steps we take to ensure our partners are well-suited for the government’s investment in AI.
